43-LB: Preliminary Evidence of the Validity of a Measure to Assess Diet Acceptability—the Advancing Care for Type 1 Diabetes and Obesity Network (ACT1ON) Study

Abstract

Diet acceptability is widely considered to be an important driver of behavioral adherence and health benefits in dietary interventions, yet there is an absence of measures with evidence of validity to assess this construct and guide decision-making. Informed by existing literature and content validity appraisal of subject matter experts, we developed and evaluated the validity of a five-item measure of diet acceptability (total score range: 5-20, total score >10 = “unacceptable”) with Likert response options (1- Strongly Agree, 2- Agree, 3- Disagree, 4- Strongly Disagree), which we administered six times during a 10.5 month Sequential, Multiple Assignment, Randomized Trial pilot that tested effects of three diets on weight and glycemic management in adults 19-30 years with type 1 diabetes ≥1 year (n=68, HbA1c <13%, BMI: 27-39). To maximize sample size (n=51), preliminary validity analyses relied predominantly on data from the first diet period (0-3 months). Greater diet acceptability according to continuous and dichotomous total and individual item scores at 3 months was associated with greater weight loss 0-3 months (e.g., total score ≤10 was associated with 2.4kg more weight loss [95% CI: 0.6, 4.2; p=0.01; n=48]). Cronbach’s α was 0.93. Exploratory factor analysis supported a 1-factor structure, explaining 99.7% of common variance. High item-total correlations (0.72 - 0.87), average inter-item correlation (0.88), and item - factor correlations (0.75 - 0.93), together with criterion validity analyses suggest item redundancy and indicate item 1 (i.e., “I am satisfied with the diet that I am currently on”) shares the strongest association with the underlying construct. Our analyses indicate robust preliminary evidence of reliability and validity of the five-item measure, while also suggesting potential adequacy of item 1 to assess the construct and minimize participant response burden in future studies that employ the measure. Disclosure A. Cristello sarteau: None. D. Igudesman: None. D. M. Maahs: Advisory Panel; Medtronic, LifeScan Diabetes Institute, MannKind Corporation, Consultant; Abbott, Research Support; Dexcom, Inc. R. E. Pratley: Other Relationship; Bayer Inc., Corcept Therapeutics, Dexcom, Inc., Gasherbrum Bio, Inc., Hanmi Pharm. Co., Ltd., Hengrui (USA) Ltd., Merck Sharp & Dohme Corp., Novo Nordisk, Pfizer Inc., Rivus Pharmaceuticals Inc., Sanofi, Scohia Pharma Inc., Sun Pharmaceutical Industries Ltd. E. J. Mayer-davis: None. Funding Clinical Trial Registration (NCT03651622)